• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.10
• /
• pp.156-165
• /
• 1998

In this paper, the hybrid algorithm for the scene change detection of MPEG-based compressed video data is proposed. There have been two methods to detect scene changes of video data compressed using algorithms such as MPEG or motion-JPEG: analyzing the compressed data directly, and analyzing from the retrieved data. The former has the advantage of taking less time, while the latter can obtain detail results at the expense of time and memory. Thus by combining each algorithm we detect cuts from compressed sequence, retrieve data for some selected region, and detect gradual scene changes. Simulation results verify the superiorities of the proposed algorithm in analyzing time and accuracy.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2018.06a
• /
• pp.322-325
• /
• 2018

영상의 관심 영역 검출은 영상처리 및 컴퓨터 비전 응용 분야에서 꾸준하게 사용되고 있는 기법이다. 특히, 근래 심층신경망 연구의 급격한 발전에 힘입어 심층신경망을 이용한 관심 영역 검출 기법에 대한 연구가 활발하게 진행되고 있다. 한편 Fully Convolutional Network(이하 FCN)은 본래 심층 예측(Dense Prediction)을 통한 의미론적 영상 분할(Semantic Segmentation)을 수행하기 위해 제안된 심층신경망 구조이다. FCN을 영상의 관심 영역 검출에 활용하여도 기존 관심 영역 검출 기법과 비교하여 충분히 좋은 성능을 발휘할 수 있다. 그러나 FCN에 사용되는 convolution 층의 수가 많고, 이에 따른 가중치(weight)의 개수도 기하급수적으로 늘어나 검출에 필요한 시간 복잡도가 매우 크다는 문제점이 있다. 따라서 본 논문에서는 기존 FCN이 가진 검출 시간 복잡도의 문제점을 convolution 층의 가중치 관점에서 해결하고자 이를 조절하여 FCN의 관심 영역 검출 속도를 향상시키는 방법을 제안한다. 적절한 convolution 층의 가중치를 조절함으로써, MSRA10K 데이터셋 환경에서 검출 정확도를 크게 저하시키지 않고도 최대 약 20.5%만큼 검출 속도를 향상시킬 수 있었다.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.3
• /
• pp.16-22
• /
• 2007

Motion Estimation time in the H.264 has a large portion of encoding time and must be improved for real time application. Most of proposed motion estimation algorithm including Sub-pixel search use the fast search algorithm to speed up motion estimation by targeting the performance of full search in the reference code. This paper proposes a novel fast sub-pixel search control algorithm for H.264 encoder by using neighbor motion vector after analyzing the encoded Motion vector of video sequence. In addition the horizontal/vertical searching method is proposed with the horizontal/vertical directionality of motion vector. And the evaluation is performed with the proposed algorithms and other reference algorithms.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.44 no.3
• /
• pp.23-32
• /
• 2007

The H.264/AVC video coding standard uses rate distortion optimization (RDO) method to improve the compression performance in the intra prediction. The complexity and computational load are increased more than previous standard by using this method, even though this standard selects the best coding mode for the current macroblock. This paper proposes a fast intra mode decision algorithm for H.264/AVC encoder based on dominant edge direction (DED). To apply the idea, this algorithm uses the approximation of discrete cosine transform (DCT) coefficient. By detecting the DED, 3 modes instead of 9 modes are chosen for RDO calculation to decide the best mode in the $4{\times}4$ luma block. As for the $16{\times}16$ luma and $8{\times}8$ chroma block, instead of 4 modes, only 2 modes are searched. Experimental results show that the computation time of the proposed algorithm is decreased to about 72% of the full search method with negligible quality loss.

• The Journal of the Acoustical Society of Korea
• /
• v.31 no.3
• /
• pp.197-204
• /
• 2012

In this thesis, studied the method to minimize the changes in frequency response and level due to the variation of the distance from the source to the microphone. selecting three microphones (omni directional, cardioid, super cardioid) which are being used generally, frequency responses were measured in accordance with the distance changes. Gotten the difference from the reference as the result of measurement, changed responses for each frequency range were compensated in comparison of the original human vocal source. In low frequency range, the low frequency boost caused by the proximity effect and decrease in accordance with the distance were compensated. The variation in mid-frequency range is comparatively small, however since the mid-range is the most important part of the human vocal signal, were compensated the mid-frequency range in comparison of the reference. The human vocal signal variation in high frequency range is extremely small and the high frequency is compensated close to the original source without difficulty. Understanding the microphone characteristics and compensations, this study showed that the response can be maintain among the change of the distance from the source to the microphone.

• Journal of Ocean Engineering and Technology
• /
• v.33 no.6
• /
• pp.597-606
• /
• 2019

Due to climate change, coastal areas are being flooded with torrential rain, typhoons, and tsunamis. In addition, non-point source pollutants (NPSs) that accumulated on the ground, streets, and buildings during the dry season are washed off by rain and stormwater runoff, which adds to the damage associated with environmental pollution, e.g., pollution that makes its way into the ocean. Recently, low impact development (LID) has been considered as a means of controlling water circulation and NPSs. In the coastal area, permeable blocks have been constructed mainly to reduce the flood damage caused by waves. Some important design factors that must be considered to ensure long-term performance are the permeability coefficient, clogging, and the efficiency of the removal of total suspended solids (TSS), but currently there are no standardized design criteria or testing techniques that are used worldwide. Herein, we analyzed the permeability coefficient and the TSS removal efficiency tendency according to the permeability area ratio with an easily-detachable, permeable block filled with calcinated yellow soils as the filter media. Our lab-scale tests indicated that, when the permeability area ratio was 25%, the reduction of the permeability coefficient after clogged was 11%, which was a significant decrease compared to other cases. Permeability persistence increased when the permeability area ratio increased from 50% to 75%. The TSS removal efficiency decreased as the permeability area ratio increased. Our pilot-scale test indicated that the TSS removal efficiency was more than 80% higher in all cases. We also found that the permeability persistence was excellent as the permeability area ratio increased, and, in actual construction, it is effective to set 5.3% of the total area as permeable area in terms of permeability and economic feasibility.

• Korean Journal of Veterinary Service
• /
• v.14 no.2
• /
• pp.110-120
• /
• 1991

28 porcine enteroviruses were isolated from 86 pig-feces of 9 swine farms located in south region, Chung-buk, from March to September 1990. Physico-chemical properties and pathogenicity of isolates were investigated. Results obtained throughout experiments are summarized as follows. According to the age, weanlings(40-90 days), sucklings(10-30 days) and adult pigs(6 months over) showed the isolation rate of 67%. 8% and 4%, respectively. By physico-chemical tests, YD-90/22, YD-90/43 and YD-90/64 strains were found to be ether, chloroform and PH stable. Nucleic acid test suggests the virus to have a DNA genome. Most of the Isolates were not evident of hemagglutinin using erythrocytes from various mammalian & avian. 22 strains among the isolates were shown CPE type I and the remainders were CPE type II. 3 strains among isolates of CPE type I strains were neutralized with high titers to serotype 2 antiserum. In the study on virus growth curve in PK-l5 cells, YD-90/22, YD-90/43 and YD-90/64 strains showed the maximum infectivity titers($10^{6.0}-l0^{6.5} TCID({50}ml$) at 4days post inoculation(PI). When 30 day-old commercial piglets were inoculated only intraoral route with the YD-90/22 strain at $10^{6.0} TCID_{50}ml,$ piglets not showed the symptoms. But piglets inoculated by intramuscle route, intraoral and intramuscle route after pretreat with dexamethasone(2.5mg /kg) for 5 days were shown the symptoms of anorexia, diarrhea, pyrexia and ataxia at 4th-6th days PI. The viral reisolation in the virus-inoculated piglets was examined from feces. The viruses were recovered intermittently from 2nd to 16th day PI and at 4th-6th day PI, all piglets excreted viruses.

• Korean Journal of Veterinary Service
• /
• v.15 no.2
• /
• pp.203-214
• /
• 1992

Livestock waste causes ground and surface water pollution, eutrophication of reservoir as well as adverse affects living environment of those who dwell nearby. In order to investigate the pollution load from livestock waste, physical and chemical characteristics of the waste were determinded in the survey of livestock farms. Once pollution load is obtained as a function of the origin units of livestock, the influence of livestock waste on the Daecheong reservoir was studied. 1. For Daecheong reservoir inflow area, the liverstock manure productions of beef cattle swine, dairy cattle and chicken were 1,135.6t/day, 480t/day, 241.3t/day, 48t/day respectively； Beef cattle was the mai or source of the pollution. Pollution loading productions due to the livestock waste around the were found to be 53.31t/day for BOD, 222.49t /day for COD, 261.99t/day for T-S, 9.64t/day for T-N, 6.54t/day for T-P 2. Bocheong stream turnd out to be the major contribution for pollution loading production to Daecheong reservoir with 10,748kg /day of BOD, 47,157kg /day of COD, 1,946kg /day of T-N, 1,271kg /day of T-p. 3. Actual pollution loadings from livestock wastes for the area of Daecheong reservoir were estimated as 1,997kg /day(BOD), 8,546kg /day(COD), 364kg /day(T-N), 243kg /day (T-P), respectively Therefore, advanced treatment for livestock waste is recommended for Daectleong reservoir inflow area to eliminate the nutrients which are major sources of eutrophication of the reservoir.

• Journal of Broadcast Engineering
• /
• v.4 no.1
• /
• pp.59-67
• /
• 1999

In this paper, we study two important topics in stereoscopic image communication system. One is a disparity estimation (DE) method to obtain the depth information of a scene at the transmitter and the other is an intermediate view reconstruction(IVR) method at the receiver. We propose a new DE method using an edge-preserving directional regularization technique. The proposed DE method smooths disparity vectors in smooth regions and preserves edges without over-smoothing problem. It provides better reconstructed stereoscopic images and improved coding efficiency than the existing regularization techniques. In addition. we propose a new IVR method using interpolation and extrapolation techniques. The proposed IVR method preserves edge regions as well as occlusion regions well. Thus. it gives better intermediate views than the existing IVR methods.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.5C
• /
• pp.464-471
• /
• 2010

The demand for high quality of multimedia applications, which far surpasses the rapid evolution of transmission and storage technologies, makes better compression coding capabilities ever increasingly more important. In order to provide enhanced video coding performance, this paper proposes an efficient competition-based motion vector coding scheme. The proposed algorithm adaptively forms the motion vector predictors based on the contexts of scene characteristics such as camera motion and nearby motion vectors, providing more efficient candidate predictors than the previous competition-based motion vector coding schemes which resort to the fixed candidates optimized by extensive simulations. Up to 200% of compression gain was observed in the experimental results for the proposed scheme applied to the motion vector selection for skip mode processing.